Methodology and Reflection: Linguistic Discourse Historiography as Digitally Supported Group Research (Methodological Cross-Sectional Project)

数字话语分析需要人工解释和算法程序相互驱动知识过程的程序。从方法论的角度来看，研究小组“有争议的话语。1990年以来作为当代史的语言史”将其整体视为理解话语史的合作方法的实验室。其目的是开发一个符合话语分析质量标准和理论框架的解释性类别数字分析工作流程。这引发了许多该领域尚未解决的程序性问题，一方面涉及实用注释的质量标准，例如注释中的分割、分类和解释深度之间的关系，另一方面涉及双重操作步骤从研究问题到注释方案再到模式识别。分项目7的目的是评价、反思和逐步优化研究小组的方法，并将其纳入协作工作流程。这是追求通过中心的研究问题：如何可以优化协作注释的过程，并制定适当的测量方法，其质量？为此，本文将对“注释作为话语语言学的一种方法”进行系统的研究。这将汇集和系统化研究小组的方法研究结果，并进行基于人工神经网络的分类器使用实验。由于这一反身性特点和独特的方法取向，该次级项目将不收集自己的研究数据，而是逐步利用其他次级项目新出现的研究数据。这种设置的一个决定性优势是，这些程序-方法问题可以与一个小组一起工作，在这个小组中，他们的话语-理论框架以及所有相关的定性和定量方法都有深刻的专业知识。

Digital discourse analysis requires procedures in which manual-hermeneutic and algorithmic procedures mutually drive the process of knowledge. From a methodological point of view, the research group “Controversial Discourses. Language History as Contemporary History since 1990” sees itself in its entirety as a laboratory of collaborative methods of an understanding discourse history. The aim is to develop a workflow for digital analyses of interpretative categories that meets the quality standards and theoretical framework of discourse analysis. This raises a number of procedural questions that are still unresolved in the field, concerning on the one hand the quality standards of pragmatic annotation, e.g. the relationship between segmentation, categorisation and depth of interpretation in annotation, and on the other hand the double operationalisation step from the research question to the annotation scheme to pattern recognition. The aim of sub-project 7 is to evaluate, reflect on and incrementally optimise the methods of the research group and their integration into a collaborative workflow. This is pursued by means of the central research question: How can processes of collaborative annotation be optimised and appropriate measurement methods of their quality be developed? To this end, a systematic study on “Annotation as a method of discourse linguistics” will be carried out. This will bring together and systematise the methodological findings of the research group and conduct experiments on the use of classifiers based on artificial neural networks. Because of this reflexive character and the exclusive methodological orientation, the sub-project will not collect its own research data, but will work incrementally with the emerging research data of the other sub-projects. A decisive advantage of this setting is that these procedural-methodological questions can be worked on with a group in which profound expertise is available for their discourse-theoretical framing as well as all relevant qualitative and quantitative methods.